← Back to peoplePerson profile
David Cahn's AI stack.
Partner at Sequoia Capital.
Tools David Cahn has publicly discussed or demonstrated using. Curated by Magpie's editorial team from podcasts, posts, and public stacks.
About
David Cahn is a partner at Sequoia Capital. He authored AI's $200B Question and AI's $600B Question, the most-cited essays on the AI capex gap. His public commentary is thesis-driven and focused on the economics of the AI buildout.
Tools they advocate for
ChatGPT4.9
DirectLLM chatbot
ChatGPT is the central reference point in Cahn's AI capex essays. His widely-cited 'AI's $600B Question' notes that 'few AI products consumers are using outside of ChatGPT' and uses OpenAI's revenue trajectory ($3.4B up from $1.6B) as the anchor data point for whether AI infrastructure spend will pay back. His thesis treats ChatGPT as the existence proof that consumer AI demand is real, while questioning whether infrastructure-tier returns can justify the build-out.
Claude4.7
DirectLLM chatbot
Cahn discusses Claude in his AI economics essays as the closest competitor to ChatGPT in the consumer AI category, and as a benchmark for tracking whether the model layer is consolidating or fragmenting. His analysis frames Claude alongside ChatGPT as the two reference points for what 'real consumer AI revenue' means versus infrastructure spend.
Perplexity3.7
DirectLLM chatbot
Cahn has discussed Perplexity in commentary on AI-search economics. He frames it as a notable third-tier consumer AI product after ChatGPT and Claude, and uses it in analysis of whether AI search can capture meaningful Google-search-revenue share. His framing is honest about the addressable market gap between agent products and chat products.
Last researched May 10, 2026
Editorial note
These tools have been mentioned by the people featured in podcasts, interviews, posts, and other public media. Source links and citations are included wherever possible. Magpie AI has no direct relationship with the people on these pages. Spot a missing tool or a wrong attribution? Suggest a correction →